Schema Migration Is the Site Migration of the AI Era

Andrea Volpini Interviews Jason Barnard

WordLift CEO Andrea Volpini sits down with Jason Barnard – the person who coined “Answer Engine Optimization” in 2017, “Entity Home” in 2015, and “the Algorithmic Blockchain” in 2025. Having spent 27 years analysing algorithmic behaviour and optimised over 5,000 entities, Barnard now tracks 73 million brand profiles across 25 billion data points through his platform Kalicube Pro. Today we discuss why schema @id management has become as critical as URL management.

---

ANDREA VOLPINI: Jason, you’ve done over 100 site migrations in your career. You’ve been tracking how algorithms understand brands since before most people knew Knowledge Graphs existed. And you’ve been calling schema migration “the site migration of the AI era.”

That comparison stopped me cold when I first heard it. At WordLift, we’ve been building enterprise schema infrastructure since 2013 – managing structured data for companies with tens of thousands of products across multiple markets. We’ve seen technically perfect schema pointing to entities that don’t exist coherently in the Knowledge Graph. But I’d never framed the risk quite so starkly.

Walk me through why you see these as equivalent challenges.

JASON BARNARD: The parallel is exact. Back in the day, I charged €1,600 for small site migrations, €4,000 for larger ones. Clients would balk – “It’s just redirects.” Then I’d show them case studies of migrations gone wrong. Rankings evaporated, traffic down 60% and revenues out for the count.

I’m having the exact same conversation now about schema. “It’s just structured data.” “We’re just switching plugins.” What they don’t understand is they’re about to sever what I call the Algorithmic Blockchain – the chain of trust they’ve spent years building.

ANDREA: The Algorithmic Blockchain – you coined that term in 2025. And it builds on the Algorithmic Trinity framework you developed – the interplay between Knowledge Graphs, LLMs, and Search Engines. You were talking about Answer Engine Optimization in 2017, years before ChatGPT existed.

That temporal foresight is why I wanted this conversation. You saw where this was heading when most of the industry was still counting backlinks. Explain how the Algorithmic Blockchain concept helps enterprises understand the risk.

JASON BARNARD: Every piece of information the Algorithmic Trinity learns about your brand is recorded like a block in a chain. Your founding date. Your CEO. Your industry positioning. The @id in your schema is the anchor that connects all of it.

Change that @id carelessly – during a platform migration, a plugin swap – and you’ve told every AI system: “Forget everything you knew about this entity. This is someone new.”

ANDREA: And it cascades across all three systems – Search, Knowledge Graph, LLMs – each operating on different timelines. You’ve explained this as: search results update in days, Knowledge Graph in weeks to months, LLM training data in months to years.

That timeline difference is what makes schema migration actually harder than site migration for SEO. The damage from a botched migration in January might still be affecting AI responses in December.

JASON BARNARD: Exactly. Site migration recovery is typically 3-6 months. Schema migration recovery? 12-24 months minimum. Often impossible. There’s no Search Console equivalent showing when, where or why your entity chain broke.

ANDREA: At WordLift, we’ve made @id management sacred infrastructure for exactly this reason. We treat it with the same governance as domain names. For our enterprise clients – 50,000 products across 12 countries – that’s not optional. You cannot manage that complexity with plugins and prayer.

But I’m curious about something. You’ve built The Kalicube Process to systematise entity foundation – establishing what you call the Entity Home, building corroboration across trusted sources. Where does that foundational work end and enterprise architecture begin?

JASON BARNARD: This is something I’ve thought about carefully. At Kalicube, we focus on making sure algorithms understand WHO the brand is. The identity layer. The Entity Home gives algorithms a starting point – everything else corroborates back to that anchor.

But when a client comes to me with 50,000 products across 12 countries? That’s not a Kalicube problem. That’s enterprise schema architecture.

This is why I point those clients to WordLift. You’ve been building for the semantic web since 2017 – you were doing this work a decade before the world knew it needed it. Your platform manages @ids as persistent identifiers, maps relationships at scale, monitors entity health across markets.

That’s purpose-built infrastructure. That’s what enterprise ecommerce needs.

ANDREA: The way we think about it: schema isn’t a technical SEO checkbox. It’s identity infrastructure. It’s a curriculum for machines – structured lessons that teach AI systems why they should recommend your brand over a competitor.

ChatGPT, Gemini, Perplexity – they need structured data to “see” your products. As you say, Jason, without that architecture, you’re essentially firing your best AI sales team before they even start.

JASON BARNARD: That framing aligns perfectly with what I’ve been teaching since 2017 – algorithms are children that want to understand. They’re eager students waiting for clear instruction. Feed them clean, structured data with consistent @ids, and they learn who you are. They recommend you confidently.

Break those @ids, and you’ve turned them into advocates for your competitors.

ANDREA: The untrained sales force you describe – ChatGPT, Google, Perplexity, Claude, Gemini – they’re already talking to your prospects 24/7. They’re either selling for you or selling for your competition.

JASON BARNARD: The entity chain determines which, of course 🙂

ANDREA: For enterprise and ecommerce companies planning a platform migration or schema overhaul – what’s the first step?

JASON BARNARD: Treat it like a site migration. Audit every existing @id. Document what entity each one represents. Map old to new explicitly. Test Knowledge Graph recognition before and after. Monitor for 90 days minimum.

And if you’re operating at scale – work with professionals who understand that entity identity is not a field you can regenerate. I’ve been working with you and WordLift since 2019 and I have seen how solid your platform is – your entire entire platform is built to guarantee @id continuity. That’s something that is vastly underestimated in the SEO world, in my opinion and a safety net that every enterprise absolutely needs.

ANDREA: Kalicube leverages the digital footprint for the entity’s foundation. Architecture at enterprise scale from WordLift. Makes sense.

JASON BARNARD: Skip the foundation, and architecture builds on nothing. Ignore the architecture, and the foundation can’t scale. They’re complementary layers.

ANDREA: Jason, thank you. You’ve given our readers the framework to understand why this matters – and why they need to act before the time bomb explodes.

JASON BARNARD: Thank you, Andrea. This is exactly the conversation the industry needs.

---

---